Objective: This paper aims to enhance exoskeleton compliance during locomotion assistance by reducing misalignment and to improve energy efficiency by overcoming the limitations posed by the bulky structure of powered rigid exoskeletons. Methods: A novel compliant knee exoskeleton, featuring a parallel elastic self-alignment mechanism, has been developed and structurally optimized. The exoskeleton uses adaptive oscillators to determine the wearer's gait phase and provides real-time assistance to the knee joint. Results: Bench tests demonstrate that the parallel elastic mechanism significantly reduces the driving torque of the knee exoskeleton. Performance evaluations reveal that, compared to a commercial orthosis, the root-mean-square of knee angle error, joint misalignment, and unexpected interaction forces are reduced by 16.5 $pm$ 11.3%, 23.3 $pm$ 4.9%, and 17.7 $pm$ 1.3%, respectively. Gait intervention experiments show reductions in average and maximum muscle activity of the knee joint by 7.6 $pm$ 4.9% and 23.2 $pm$ 5.7%, respectively. Additionally, the exoskeleton decreases negative work performed by the knee joint and the total lower limb by 22.7% and 8.6%, respectively. Conclusion: The parallel elastic self-alignment mechanism effectively mitigates joint misalignment, while the parallel springs offer partial gravity compensation, thereby enhancing both the energy efficiency and locomotion assistance of the exoskeleton. Significance: The parallel elastic self-alignment mechanism effectively addresses both misalignment and energy efficiency challenges in powered exoskeletons, providing valuable insights for future design improvements.
{"title":"Parallel Elastic Self-Alignment Mechanism Enhances Energy Efficiency and Reduces Misalignment in a Powered Knee Exoskeleton","authors":"Jing Zhang;Aibin Zhu;Jiyuan Song;Bingsheng Bao;Yuxiang Su;Peng Xu;Chunli Zheng;Lei Shi;Xiaodong Zhang;Xiao Li","doi":"10.1109/TBME.2024.3461880","DOIUrl":"10.1109/TBME.2024.3461880","url":null,"abstract":"<italic>Objective</i>: This paper aims to enhance exoskeleton compliance during locomotion assistance by reducing misalignment and to improve energy efficiency by overcoming the limitations posed by the bulky structure of powered rigid exoskeletons. <italic>Methods</i>: A novel compliant knee exoskeleton, featuring a parallel elastic self-alignment mechanism, has been developed and structurally optimized. The exoskeleton uses adaptive oscillators to determine the wearer's gait phase and provides real-time assistance to the knee joint. <italic>Results</i>: Bench tests demonstrate that the parallel elastic mechanism significantly reduces the driving torque of the knee exoskeleton. Performance evaluations reveal that, compared to a commercial orthosis, the root-mean-square of knee angle error, joint misalignment, and unexpected interaction forces are reduced by 16.5 <inline-formula><tex-math>$pm$</tex-math></inline-formula> 11.3%, 23.3 <inline-formula><tex-math>$pm$</tex-math></inline-formula> 4.9%, and 17.7 <inline-formula><tex-math>$pm$</tex-math></inline-formula> 1.3%, respectively. Gait intervention experiments show reductions in average and maximum muscle activity of the knee joint by 7.6 <inline-formula><tex-math>$pm$</tex-math></inline-formula> 4.9% and 23.2 <inline-formula><tex-math>$pm$</tex-math></inline-formula> 5.7%, respectively. Additionally, the exoskeleton decreases negative work performed by the knee joint and the total lower limb by 22.7% and 8.6%, respectively. <italic>Conclusion</i>: The parallel elastic self-alignment mechanism effectively mitigates joint misalignment, while the parallel springs offer partial gravity compensation, thereby enhancing both the energy efficiency and locomotion assistance of the exoskeleton. <italic>Significance</i>: The parallel elastic self-alignment mechanism effectively addresses both misalignment and energy efficiency challenges in powered exoskeletons, providing valuable insights for future design improvements.","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"72 2","pages":"528-539"},"PeriodicalIF":4.4,"publicationDate":"2024-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142265008","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-16DOI: 10.1109/TBME.2024.3460814
Kyeongho Eom;Han-Sol Lee;Minju Park;Seung Min Yang;Jong Chan Choe;Suk-Won Hwang;Young-Woo Suh;Hyung-Min Lee
Paralysis of the extraocular muscles can lead to complications such as strabismus, diplopia, and loss of stereopsis. Current surgical treatments aim to mitigate these issues by resecting the paralyzed muscle or transposing the other recti muscles to the paralyzed muscle, but they do not fully improve the patient's quality of life. Electrical stimulation shows promise, while requiring further in vivo experiments and research on various stimulation parameters. In this study, we conducted experiments on rabbits to stimulate the superior rectus (SR) muscles using different parameters and stimulation waveforms. To provide various types of electrical stimulation, we developed the ocular muscle stimulation systems capable of both current controlled stimulation (CCS) and high-frequency stimulation (HFS), along with the chip that enables energy-efficient and safe switched-capacitor stimulation (SCS). We also developed electrodes for easy implantation and employed safe and efficient stimulation methods including CCS, SCS, and HFS. The in vivo animal experiments on normal and paralyzed SR muscles of rabbits showed that eyeball abduction angles were proportional to the current and pulse width of the stimulation. With the decaying exponential stimuli of the SCS system, eyeball abductions were 2.58× and 5.65× larger for normal and paralyzed muscles, respectively, compared to the rectangular stimulus of CCS. HFS achieved 0.92× and 0.26× abduction for normal and paralyzed muscles, respectively, with half energy compared to CCS. In addition, the continuous changes in eyeball abduction angle in response to varying stimulation intensity over time were observed.
{"title":"Development of Ocular Muscle Stimulation Systems and Optimization of Electrical Stimulus Parameters for Paralytic Strabismus Treatment","authors":"Kyeongho Eom;Han-Sol Lee;Minju Park;Seung Min Yang;Jong Chan Choe;Suk-Won Hwang;Young-Woo Suh;Hyung-Min Lee","doi":"10.1109/TBME.2024.3460814","DOIUrl":"10.1109/TBME.2024.3460814","url":null,"abstract":"Paralysis of the extraocular muscles can lead to complications such as strabismus, diplopia, and loss of stereopsis. Current surgical treatments aim to mitigate these issues by resecting the paralyzed muscle or transposing the other recti muscles to the paralyzed muscle, but they do not fully improve the patient's quality of life. Electrical stimulation shows promise, while requiring further in vivo experiments and research on various stimulation parameters. In this study, we conducted experiments on rabbits to stimulate the superior rectus (SR) muscles using different parameters and stimulation waveforms. To provide various types of electrical stimulation, we developed the ocular muscle stimulation systems capable of both current controlled stimulation (CCS) and high-frequency stimulation (HFS), along with the chip that enables energy-efficient and safe switched-capacitor stimulation (SCS). We also developed electrodes for easy implantation and employed safe and efficient stimulation methods including CCS, SCS, and HFS. The in vivo animal experiments on normal and paralyzed SR muscles of rabbits showed that eyeball abduction angles were proportional to the current and pulse width of the stimulation. With the decaying exponential stimuli of the SCS system, eyeball abductions were 2.58× and 5.65× larger for normal and paralyzed muscles, respectively, compared to the rectangular stimulus of CCS. HFS achieved 0.92× and 0.26× abduction for normal and paralyzed muscles, respectively, with half energy compared to CCS. In addition, the continuous changes in eyeball abduction angle in response to varying stimulation intensity over time were observed.","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"72 2","pages":"515-527"},"PeriodicalIF":4.4,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142286070","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-11DOI: 10.1109/TBME.2024.3458060
Di Chen;Zhiqing Song;Yang Du;Sicong Chen;Xin Zhang;Yuanqing Li;Qiyun Huang
Objective: In this study, we aimed to investigate whether and how the aperiodic component in electroencephalograms affects different quantitative processes of steady-state visually evoked potentials and the performance of corresponding brain-computer interfaces. Methods: We applied the Fitting Oscillations & One-Over-F method to parameterize power spectra as a combination of periodic oscillations and an aperiodic component. Electroencephalographic responses and system performance were measured and compared using four prevailing methods: power spectral density analysis, canonical correlation analysis, filter bank canonical correlation analysis and the state-of-the-art method, task discriminant component analysis. Results: We found that controlling for the aperiodic component prominently downgraded the performance of brain-computer interfaces measured by canonical correlation analysis (94.9% to 82.8%), filter bank canonical correlation analysis (94.1% to 87.6%), and task discriminant component analysis (96.5% to 70.3%). However, it had almost no effect on that measured by power spectral density analysis (80.4% to 78.7%). This was accompanied by a differential aperiodic impact between power spectral density analysis and the other three methods on the differentiation of the target and non-target stimuli. Conclusion: The aperiodic component distinctly impacts the quantification of steady-state visually evoked potentials and the performance of corresponding brain-computer interfaces. Significance: Our work underscores the significance of taking into account the dynamic nature of aperiodic activities in research related to the quantification of steady-state visually evoked potentials.
目的：本研究旨在探讨脑电图非周期成分是否以及如何影响稳态视觉诱发电位的不同定量过程和相应的脑机接口的性能。方法：采用拟合振荡和1 - over - f方法将功率谱作为周期振荡和非周期分量的组合参数化。采用功率谱密度分析、典型相关分析、滤波器组典型相关分析和最先进的任务判别成分分析四种常用方法测量和比较脑电反应和系统性能。结果：我们发现，控制非周期成分显著降低了典型相关分析（94.9%至82.8%）、滤波器组典型相关分析（94.1%至87.6%）和任务判别成分分析（96.5%至70.3%）测量的脑机接口性能。然而，它对功率谱密度分析的测量结果几乎没有影响（80.4% ~ 78.7%）。与此同时，功率谱密度分析与其他三种方法对目标和非目标刺激的区分具有不同的非周期影响。结论：非周期成分明显影响稳态视觉诱发电位的定量及相应的脑机接口性能。意义：我们的工作强调了在与稳态视觉诱发电位量化相关的研究中考虑非周期性活动的动态性质的重要性。
{"title":"Aperiodic Component Analysis in Quantification of Steady-State Visually Evoked Potentials","authors":"Di Chen;Zhiqing Song;Yang Du;Sicong Chen;Xin Zhang;Yuanqing Li;Qiyun Huang","doi":"10.1109/TBME.2024.3458060","DOIUrl":"10.1109/TBME.2024.3458060","url":null,"abstract":"<italic>Objective:</i> In this study, we aimed to investigate whether and how the aperiodic component in electroencephalograms affects different quantitative processes of steady-state visually evoked potentials and the performance of corresponding brain-computer interfaces. <italic>Methods:</i> We applied the Fitting Oscillations & One-Over-F method to parameterize power spectra as a combination of periodic oscillations and an aperiodic component. Electroencephalographic responses and system performance were measured and compared using four prevailing methods: power spectral density analysis, canonical correlation analysis, filter bank canonical correlation analysis and the state-of-the-art method, task discriminant component analysis. <italic>Results:</i> We found that controlling for the aperiodic component prominently downgraded the performance of brain-computer interfaces measured by canonical correlation analysis (94.9% to 82.8%), filter bank canonical correlation analysis (94.1% to 87.6%), and task discriminant component analysis (96.5% to 70.3%). However, it had almost no effect on that measured by power spectral density analysis (80.4% to 78.7%). This was accompanied by a differential aperiodic impact between power spectral density analysis and the other three methods on the differentiation of the target and non-target stimuli. <italic>Conclusion:</i> The aperiodic component distinctly impacts the quantification of steady-state visually evoked potentials and the performance of corresponding brain-computer interfaces. <italic>Significance:</i> Our work underscores the significance of taking into account the dynamic nature of aperiodic activities in research related to the quantification of steady-state visually evoked potentials.","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"72 2","pages":"468-479"},"PeriodicalIF":4.4,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10675437","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142175216","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-10DOI: 10.1109/TBME.2024.3454067
S. Bhattacharya;F. Santucci;M. Jankovic;T. Huang;J. Basu;P. Tan;E. Schena;N. Lu
Goal: We present a wireless, lightweight, stretchable, and chest-conformable sensor, known as the chest e-tattoo, coupled with an advanced signal processing framework to accurately identify various cardiac events, and thereby extract cardiac time intervals (CTIs) even during body motion. Methods: We developed a wireless chest e-tattoo featuring synchronous electrocardiography (ECG) and seismocardiography (SCG), with SCG capturing chest vibrations to complement ECG. Motion artifacts often compromise the efficacy of SCG, but the e-tattoo's slim, stretchy design allows strategic placement near the xiphoid process for improved signal quality. Nine participants were monitored during walking and cycling. To accurately extract CTIs, we implemented a multistage signal processing framework, named the FAD framework, combining adaptive Normalized Least Mean Squares (NMLS) filtering, ensemble averaging, and Empirical Mode Decomposition (EMD). Results: Key CTIs, especially left ventricular ejection time (LVET), were successfully extracted by our hardware-software system and showed strong agreement with those reported by an FDA-cleared bedside monitor even during substantial movements. The pre-ejection period (PEP) measured by the e-tattoo also aligned with previous findings. Conclusion: The bimodal chest e-tattoo combined with the FAD framework enables reliable CTI measurements during various activities. Significance: Managing cardiovascular disease at home necessitates continuous monitoring, which has been challenging with wearables due to signal sensitivity to motion. Accurately extracting cardiac events from synchronous SCG and ECG during motion can significantly enhance heart stress response quantification, offering a more comprehensive cardiac health assessment than ECG alone and marking a significant advancement in ambulatory cardiovascular monitoring capabilities.
{"title":"Cardiac Time Intervals Under Motion Using Bimodal Chest E-Tattoos and Multistage Processing","authors":"S. Bhattacharya;F. Santucci;M. Jankovic;T. Huang;J. Basu;P. Tan;E. Schena;N. Lu","doi":"10.1109/TBME.2024.3454067","DOIUrl":"10.1109/TBME.2024.3454067","url":null,"abstract":"<italic>Goal:</i> We present a wireless, lightweight, stretchable, and chest-conformable sensor, known as the chest e-tattoo, coupled with an advanced signal processing framework to accurately identify various cardiac events, and thereby extract cardiac time intervals (CTIs) even during body motion. <italic>Methods:</i> We developed a wireless chest e-tattoo featuring synchronous electrocardiography (ECG) and seismocardiography (SCG), with SCG capturing chest vibrations to complement ECG. Motion artifacts often compromise the efficacy of SCG, but the e-tattoo's slim, stretchy design allows strategic placement near the xiphoid process for improved signal quality. Nine participants were monitored during walking and cycling. To accurately extract CTIs, we implemented a multistage signal processing framework, named the FAD framework, combining adaptive Normalized Least Mean Squares (NMLS) filtering, ensemble averaging, and Empirical Mode Decomposition (EMD). <italic>Results:</i> Key CTIs, especially left ventricular ejection time (LVET), were successfully extracted by our hardware-software system and showed strong agreement with those reported by an FDA-cleared bedside monitor even during substantial movements. The pre-ejection period (PEP) measured by the e-tattoo also aligned with previous findings. <italic>Conclusion:</i> The bimodal chest e-tattoo combined with the FAD framework enables reliable CTI measurements during various activities. <italic>Significance:</i> Managing cardiovascular disease at home necessitates continuous monitoring, which has been challenging with wearables due to signal sensitivity to motion. Accurately extracting cardiac events from synchronous SCG and ECG during motion can significantly enhance heart stress response quantification, offering a more comprehensive cardiac health assessment than ECG alone and marking a significant advancement in ambulatory cardiovascular monitoring capabilities.","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"72 1","pages":"413-424"},"PeriodicalIF":4.4,"publicationDate":"2024-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142175223","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-10DOI: 10.1109/TBME.2024.3458177
Tala Abdallah;Nisrine Jrad;Sally El Hajjar;Fahed Abdallah;Anne Humeau-Heurtier;Eliane El Howayek;Patrick Van Bogaert
Epilepsy is a neurological disorder characterized by recurrent epileptic seizures, which are often unpredictable and increase mortality and morbidity risks. Objective: The objective of this study is to address the challenges of EEG-based epileptic seizure detection by introducing a novel methodology, Deep Embedded Gaussian Mixture (DEGM). Methods: The DEGM method begins with a deep autoencoder (DAE) for embedding the input EEG data, followed by Singular Value Decomposition (SVD) to enhance the representational quality of the embedding while achieving dimensionality reduction. A Gaussian Mixture Model (GMM) is then employed for clustering purposes. Unlike conventional supervised machine learning and deep learning techniques, DEGM leverages deep clustering (DC) algorithms for more effective seizure detection. Results: Empirical results from two real-world epileptic datasets demonstrate the notable performance of DEGM. The method's effectiveness is particularly remarkable given the substantial size of the datasets, showcasing its ability to handle large-scale EEG data efficiently. Conclusion: In conclusion, the DEGM methodology provides a novel and effective approach for EEG-based epileptic seizure detection, addressing key challenges such as data variability and artifact contamination. Significance: By combining deep autoencoders, SVD, and GMM, DEGM achieves superior clustering performance compared to existing methods, representing a significant advancement in biomedical research and clinical applications for epilepsy. Its robust performance on large datasets underscores its potential for improving seizure detection accuracy, ultimately contributing to better patient outcomes.
癫痫是一种以复发性癫痫发作为特征的神经系统疾病，其往往不可预测，并增加死亡率和发病率风险。目的：本研究的目的是通过引入一种新的方法，深度嵌入高斯混合（DEGM）来解决基于脑电图的癫痫发作检测的挑战。方法：DEGM方法首先使用深度自编码器（deep autoencoder， DAE）对输入的脑电数据进行嵌入，然后使用奇异值分解（Singular Value Decomposition， SVD）在实现降维的同时提高嵌入的表征质量。然后采用高斯混合模型（GMM）进行聚类。与传统的监督机器学习和深度学习技术不同，DEGM利用深度聚类（DC）算法进行更有效的癫痫检测。结果：来自两个真实世界癫痫数据集的实证结果证明了DEGM的显著性能。考虑到数据集的庞大规模，该方法的有效性尤其显着，显示了其有效处理大规模脑电数据的能力。结论：总之，DEGM方法为基于脑电图的癫痫发作检测提供了一种新颖有效的方法，解决了数据变异性和伪影污染等关键挑战。意义：DEGM将深度自编码器、SVD和GMM相结合，取得了优于现有方法的聚类性能，在癫痫的生物医学研究和临床应用方面取得了重大进展。它在大型数据集上的强大性能强调了其提高癫痫检测准确性的潜力，最终有助于改善患者的预后。
{"title":"Deep Clustering for Epileptic Seizure Detection","authors":"Tala Abdallah;Nisrine Jrad;Sally El Hajjar;Fahed Abdallah;Anne Humeau-Heurtier;Eliane El Howayek;Patrick Van Bogaert","doi":"10.1109/TBME.2024.3458177","DOIUrl":"10.1109/TBME.2024.3458177","url":null,"abstract":"Epilepsy is a neurological disorder characterized by recurrent epileptic seizures, which are often unpredictable and increase mortality and morbidity risks. <italic>Objective:</i> The objective of this study is to address the challenges of EEG-based epileptic seizure detection by introducing a novel methodology, Deep Embedded Gaussian Mixture (DEGM). <italic>Methods:</i> The DEGM method begins with a deep autoencoder (DAE) for embedding the input EEG data, followed by Singular Value Decomposition (SVD) to enhance the representational quality of the embedding while achieving dimensionality reduction. A Gaussian Mixture Model (GMM) is then employed for clustering purposes. Unlike conventional supervised machine learning and deep learning techniques, DEGM leverages deep clustering (DC) algorithms for more effective seizure detection. <italic>Results:</i> Empirical results from two real-world epileptic datasets demonstrate the notable performance of DEGM. The method's effectiveness is particularly remarkable given the substantial size of the datasets, showcasing its ability to handle large-scale EEG data efficiently. <italic>Conclusion:</i> In conclusion, the DEGM methodology provides a novel and effective approach for EEG-based epileptic seizure detection, addressing key challenges such as data variability and artifact contamination. <italic>Significance:</i> By combining deep autoencoders, SVD, and GMM, DEGM achieves superior clustering performance compared to existing methods, representing a significant advancement in biomedical research and clinical applications for epilepsy. Its robust performance on large datasets underscores its potential for improving seizure detection accuracy, ultimately contributing to better patient outcomes.","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"72 2","pages":"480-492"},"PeriodicalIF":4.4,"publicationDate":"2024-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142175219","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Objective: The application of transfer learning, specifically pre-training and fine-tuning, in steady-state visual evoked potential (SSVEP)-based brain-computer interfaces (BCIs) has been demonstrated to effectively improve the classification performance of deep learning methods with limited calibration data. However, effectively learning task-related knowledge from source domains during the pre-training phase remains challenging. To address this issue, this study proposes an effective data augmentation method called Reconstruction of Channel Correlation (RCC) to optimize the utilization of the source domain data. Methods: Concretely, RCC reconstructs training samples using probabilistically mixed eigenvector matrices derived from covariance matrices across source domains. This process manipulates the channel correlation of training samples, implicitly creating novel synthesized domains. By increasing the diversity of source domains, RCC aims to enhance the domain generalizability of the pre-trained model. The effectiveness of RCC is validated through subject-independent and subject-adaptive classification experiments. Results: The results of subject-independent classification demonstrate that RCC significantly improves the classification performance of the pre-trained model on unseen target subjects. Moreover, when compared to the fine-tuning process using the RCC-absent pre-trained model, the fine-tuning process using the RCC-enhanced pre-trained model yields significantly improved performance in the subject-adaptive classification. Conclusion: RCC proves to enhance the performance of transfer learning by optimizing the utilization of the source domain data. Significance: The RCC-enhanced transfer learning has the potential to facilitate the practical implementation of SSVEP-BCIs in real-world scenarios.
{"title":"Enhancing Domain Diversity of Transfer Learning-Based SSVEP-BCIs by the Reconstruction of Channel Correlation","authors":"Wenlong Ding;Aiping Liu;Chengjuan Xie;Kai Wang;Xun Chen","doi":"10.1109/TBME.2024.3458389","DOIUrl":"10.1109/TBME.2024.3458389","url":null,"abstract":"<italic>Objective</i>: The application of transfer learning, specifically pre-training and fine-tuning, in steady-state visual evoked potential (SSVEP)-based brain-computer interfaces (BCIs) has been demonstrated to effectively improve the classification performance of deep learning methods with limited calibration data. However, effectively learning task-related knowledge from source domains during the pre-training phase remains challenging. To address this issue, this study proposes an effective data augmentation method called Reconstruction of Channel Correlation (RCC) to optimize the utilization of the source domain data. <italic>Methods</i>: Concretely, RCC reconstructs training samples using probabilistically mixed eigenvector matrices derived from covariance matrices across source domains. This process manipulates the channel correlation of training samples, implicitly creating novel synthesized domains. By increasing the diversity of source domains, RCC aims to enhance the domain generalizability of the pre-trained model. The effectiveness of RCC is validated through subject-independent and subject-adaptive classification experiments. <italic>Results</i>: The results of subject-independent classification demonstrate that RCC significantly improves the classification performance of the pre-trained model on unseen target subjects. Moreover, when compared to the fine-tuning process using the RCC-absent pre-trained model, the fine-tuning process using the RCC-enhanced pre-trained model yields significantly improved performance in the subject-adaptive classification. <italic>Conclusion</i>: RCC proves to enhance the performance of transfer learning by optimizing the utilization of the source domain data. <italic>Significance</i>: The RCC-enhanced transfer learning has the potential to facilitate the practical implementation of SSVEP-BCIs in real-world scenarios.","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"72 2","pages":"503-514"},"PeriodicalIF":4.4,"publicationDate":"2024-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142175218","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Objective: Controlling and coordinating finger force is crucial for performing everyday tasks and maintaining functional independence. Aging naturally weakens neural, muscular, and musculoskeletal systems, leading to compromised hand motor function. This decline reduces cortical activity, finger force control and coordination in older adults. Objective: To investigate independently the EEG band power and finger force coordination in older individuals and compare the results with young healthy adults. Methods: Twenty healthy young adults aged 20–30 (26.96 ± 2.68) and fourteen older adults aged 58–72 (62.57 ± 3.58) participated in this study. Participants held the instrumented handle gently for five seconds then lifted and held it for an additional five seconds in the two conditions: fixed (thumb platform secured) and free condition (thumb platform may slide on slider). Results: In the older individuals there was no difference observed in the finger force synergy indices, and EEG beta band power between the two task conditions. However, in the young group synergy indices and EEG beta band power were less in free condition compared to fixed condition. Additionally, in the fixed condition, older adults showed a reduced synergy indices and reduced EEG beta band power than the young adults. Conclusion: Older participants exhibited consistent synergy indices and beta band power across conditions, while young adults adjusted strategies based on tasks. Task-dependent finger force synergy indices were observed in young adults, contrasting with older individuals Additionally, it is suggested that EEG band power and synergy indices may be related, indicating potential associations between EEG activity and finger force coordination.
{"title":"Older Individuals Do Not Show Task Specific Variations in EEG Band Power and Finger Force Coordination","authors":"Balasubramanian Eswari;Sivakumar Balasubramanian;Varadhan SKM","doi":"10.1109/TBME.2024.3435480","DOIUrl":"10.1109/TBME.2024.3435480","url":null,"abstract":"<italic>Objective:</i> Controlling and coordinating finger force is crucial for performing everyday tasks and maintaining functional independence. Aging naturally weakens neural, muscular, and musculoskeletal systems, leading to compromised hand motor function. This decline reduces cortical activity, finger force control and coordination in older adults. Objective: To investigate independently the EEG band power and finger force coordination in older individuals and compare the results with young healthy adults. Methods: Twenty healthy young adults aged 20–30 (26.96 ± 2.68) and fourteen older adults aged 58–72 (62.57 ± 3.58) participated in this study. Participants held the instrumented handle gently for five seconds then lifted and held it for an additional five seconds in the two conditions: fixed (thumb platform secured) and free condition (thumb platform may slide on slider). Results: In the older individuals there was no difference observed in the finger force synergy indices, and EEG beta band power between the two task conditions. However, in the young group synergy indices and EEG beta band power were less in free condition compared to fixed condition. Additionally, in the fixed condition, older adults showed a reduced synergy indices and reduced EEG beta band power than the young adults. Conclusion: Older participants exhibited consistent synergy indices and beta band power across conditions, while young adults adjusted strategies based on tasks. Task-dependent finger force synergy indices were observed in young adults, contrasting with older individuals Additionally, it is suggested that EEG band power and synergy indices may be related, indicating potential associations between EEG activity and finger force coordination.","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"72 1","pages":"4-13"},"PeriodicalIF":4.4,"publicationDate":"2024-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142175224","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Objective: Surface electromyography (sEMG) driven musculoskeletal models are promising to be applied in the field of human-computer interaction. However, due to the individual-specific physiological characteristics, generic models often fail to provide accurate motion estimation. This study optimized the general model to build a personalized model and improve the accuracy of motion estimation. Methods: Inspired by the coupling effect of wrist/hand movement, a hierarchical optimization approach for personalizing musculoskeletal models (HOPE-MM) is proposed, which aligns with the physiological characteristics of the human wrist and hand. To verify the effectiveness of personalized musculoskeletal model, single joint motions and simultaneous joint motions are estimated. In addition, Sobol sensitivity analysis is conducted to identify the key parameters of musculoskeletal model, providing guidance for model simplification. Results: The mean pearson correlation coefficient between the predicted joint angles and the measured joint angles are 0.95 $pm$ 0.03 and 0.93 $pm$ 0.01 for simultaneous wrist and metacarpophalangeal (MCP) joint movements, respectively, which have a significant improvement compared with the state-of-the-art works. By optimizing only the key parameters including tendon slack length, maximal isometric force and optimal fiber length, the performances of simplified model are comparable to the full-parameter model. Conclusion: These results provide insights into the effects of muscle-tendon parameters on musculoskeletal model, and musculoskeletal models personalized using hierarchical optimization methods can improve the accuracy of motion estimates. Significance: These findings facilitate the clinical application of musculoskeletal models in rehabilitation and robotic control.
{"title":"Hierarchical Optimization for Personalized Hand and Wrist Musculoskeletal Modeling and Motion Estimation","authors":"Lijun Han;Long Cheng;Houcheng Li;Yongxiang Zou;Shijie Qin;Ming Zhou","doi":"10.1109/TBME.2024.3456235","DOIUrl":"10.1109/TBME.2024.3456235","url":null,"abstract":"<italic>Objective:</i> Surface electromyography (sEMG) driven musculoskeletal models are promising to be applied in the field of human-computer interaction. However, due to the individual-specific physiological characteristics, generic models often fail to provide accurate motion estimation. This study optimized the general model to build a personalized model and improve the accuracy of motion estimation. <italic>Methods:</i> Inspired by the coupling effect of wrist/hand movement, a hierarchical optimization approach for personalizing musculoskeletal models (HOPE-MM) is proposed, which aligns with the physiological characteristics of the human wrist and hand. To verify the effectiveness of personalized musculoskeletal model, single joint motions and simultaneous joint motions are estimated. In addition, Sobol sensitivity analysis is conducted to identify the key parameters of musculoskeletal model, providing guidance for model simplification. <italic>Results:</i> The mean pearson correlation coefficient between the predicted joint angles and the measured joint angles are 0.95 <inline-formula><tex-math>$pm$</tex-math></inline-formula> 0.03 and 0.93 <inline-formula><tex-math>$pm$</tex-math></inline-formula> 0.01 for simultaneous wrist and metacarpophalangeal (MCP) joint movements, respectively, which have a significant improvement compared with the state-of-the-art works. By optimizing only the key parameters including tendon slack length, maximal isometric force and optimal fiber length, the performances of simplified model are comparable to the full-parameter model. <italic>Conclusion:</i> These results provide insights into the effects of muscle-tendon parameters on musculoskeletal model, and musculoskeletal models personalized using hierarchical optimization methods can improve the accuracy of motion estimates. <italic>Significance:</i> These findings facilitate the clinical application of musculoskeletal models in rehabilitation and robotic control.","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"72 1","pages":"454-465"},"PeriodicalIF":4.4,"publicationDate":"2024-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142175225","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-06DOI: 10.1109/TBME.2024.3455270
Daniel Comadurán Márquez;Sarah J. Anderson;Kent G. Hecker;Kartikeya Murari
Objective: This work describes current-mode electroencephalography amplifiers to record the electrical activity of the tangentially oriented cells. Electroencephalography (EEG) measures the summed electrical activity from pyramidal cells in the brain by using non-invasive electrodes placed on the scalp. Traditional, voltage-based measurements are done with differential amplifiers. Depending on the location of the electrodes used for the differential measurement, EEG can estimate electrical activity from radially (common or average reference) or tangentially (bipolar derivation) oriented neurons. A limitation of the bipolar derivation is that when the electrodes are too close together, the conductive solution used to improve electrode-skin impedance can short-circuit the electrodes. Magnetoencephalography (MEG) also enables measurements from tangentially oriented cells without concerns about short-circuiting the electrodes. However, MEG is a more expensive, and a less available technology. Measuring from both radial and tangential cells can improve the resolution to localize the origin of brain activity; this could be extremely useful for diagnoses and treatment of several neurological disorders. Methods: Circuit design from previous implementations was improved and the device was compared to a voltage-based (vEEG) amplifier in a saline phantom and in humans with a steady state visually evoked potentials paradigm. Results: The current-based (cEEG) amplifier satisfied suggested electrical parameters for EEG amplifiers and exhibited higher sensitivity to tangential dipoles in the phantom study. It measured brain activity using the same scalp electrodes as vEEG amplifiers with comparable performance. Conclusion: current-based EEG amplifiers can be comparable to traditional voltage-based amplifiers and offer complementary information.
{"title":"A Current-Based EEG Amplifier and Validation With a Saline Phantom and an SSVEP Paradigm","authors":"Daniel Comadurán Márquez;Sarah J. Anderson;Kent G. Hecker;Kartikeya Murari","doi":"10.1109/TBME.2024.3455270","DOIUrl":"10.1109/TBME.2024.3455270","url":null,"abstract":"<italic>Objective:</i> This work describes current-mode electroencephalography amplifiers to record the electrical activity of the tangentially oriented cells. Electroencephalography (EEG) measures the summed electrical activity from pyramidal cells in the brain by using non-invasive electrodes placed on the scalp. Traditional, voltage-based measurements are done with differential amplifiers. Depending on the location of the electrodes used for the differential measurement, EEG can estimate electrical activity from radially (common or average reference) or tangentially (bipolar derivation) oriented neurons. A limitation of the bipolar derivation is that when the electrodes are too close together, the conductive solution used to improve electrode-skin impedance can short-circuit the electrodes. Magnetoencephalography (MEG) also enables measurements from tangentially oriented cells without concerns about short-circuiting the electrodes. However, MEG is a more expensive, and a less available technology. Measuring from both radial and tangential cells can improve the resolution to localize the origin of brain activity; this could be extremely useful for diagnoses and treatment of several neurological disorders. <italic>Methods:</i> Circuit design from previous implementations was improved and the device was compared to a voltage-based (vEEG) amplifier in a saline phantom and in humans with a steady state visually evoked potentials paradigm. <italic>Results:</i> The current-based (cEEG) amplifier satisfied suggested electrical parameters for EEG amplifiers and exhibited higher sensitivity to tangential dipoles in the phantom study. It measured brain activity using the same scalp electrodes as vEEG amplifiers with comparable performance. <italic>Conclusion:</i> current-based EEG amplifiers can be comparable to traditional voltage-based amplifiers and offer complementary information.","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"72 1","pages":"445-453"},"PeriodicalIF":4.4,"publicationDate":"2024-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142142972","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-05DOI: 10.1109/TBME.2024.3438272
Amir Esrafilian;Shekhar S. Chandra;Anthony A. Gatti;Mikko J. Nissi;Anne-Mari Mustonen;Laura Säisänen;Jusa Reijonen;Petteri Nieminen;Petro Julkunen;Juha Töyräs;David J. Saxby;David G. Lloyd;Rami K. Korhonen
Objective: To develop and assess an automatic and robust knee musculoskeletal finite element (MSK-FE) modeling pipeline. Methods: Magnetic resonance images (MRIs) were used to train nnU-Net networks for auto-segmentation of knee bones (femur, tibia, patella, and fibula), cartilages (femur, tibia, and patella), menisci, and major knee ligaments. Two different MRI sequences were used to broaden applicability. Next, we created MSK-FE models of an unseen dataset using two MSK-FE modeling pipelines: template-based and auto-meshing. MSK models had personalized knee geometries with multi-degree-of-freedom elastic foundation contacts. FE models used fibril-reinforced poroviscoelastic swelling material models for cartilages and menisci. Results: Volumes of knee bones, cartilages, and menisci did not significantly differ (p>0.05) across MRI sequences. MSK models estimated secondary knee kinematics during passive knee flexion tests consistent with in vivo and simulation-based values from the literature. Between the template-based and auto-meshing FE models, estimated cartilage mechanics often differed significantly (p<0.05),>Conclusion: The template-based modeling provided a more rapid and robust tool than the auto-meshing approach, while the estimated knee biomechanics were comparable. Nonetheless, the auto-meshing approach might provide more accurate estimates in subjects with distinct knee irregularities, e.g., cartilage lesions. Significance: The MSK-FE modeling tool provides a rapid, easy-to-use, and robust approach for investigating task- and person-specific mechanical responses of the knee cartilage and menisci, holding significant promise, e.g., in personalized rehabilitation planning.
:目的：开发并评估自动、稳健的膝关节肌肉骨骼有限元（MSK-FE）建模管道：使用磁共振成像（MRI）训练 nnU-Net 网络，以自动分割膝关节骨骼（股骨、胫骨、髌骨和腓骨）、软骨（股骨、胫骨和髌骨）、半月板和主要膝关节韧带。为了扩大适用范围，我们使用了两种不同的磁共振成像序列。接下来，我们使用两种 MSK-FE 建模流水线：基于模板和自动匹配，创建了未见数据集的 MSK-FE 模型。MSK 模型具有个性化的膝关节几何形状和多自由度弹性地基接触。软骨和半月板的 FE 模型采用纤维增强的多孔膨胀弹性材料模型：结果：不同核磁共振成像序列中膝关节骨骼、软骨和半月板的体积差异不大（P>0.05）。在膝关节被动屈曲试验中，MSK 模型估计的膝关节次要运动学特性与文献中的活体和模拟值一致。在基于模板的模型和自动套合 FE 模型之间，估计的软骨力学往往存在显著差异（p 结论：与自动镶嵌法相比，基于模板的建模方法提供了一种更快速、更稳健的工具，而估算的膝关节生物力学结果却不相上下。不过，对于膝关节明显不规则（如软骨损伤）的受试者，自动镶嵌法可能会提供更准确的估计：MSK-FE建模工具提供了一种快速、易用且稳健的方法，用于研究任务和个人特定的膝关节软骨和半月板机械响应，在个性化康复规划等方面具有重要前景。
{"title":"An Automated and Robust Tool for Musculoskeletal and Finite Element Modeling of the Knee Joint","authors":"Amir Esrafilian;Shekhar S. Chandra;Anthony A. Gatti;Mikko J. Nissi;Anne-Mari Mustonen;Laura Säisänen;Jusa Reijonen;Petteri Nieminen;Petro Julkunen;Juha Töyräs;David J. Saxby;David G. Lloyd;Rami K. Korhonen","doi":"10.1109/TBME.2024.3438272","DOIUrl":"10.1109/TBME.2024.3438272","url":null,"abstract":"<italic>Objective:</i> To develop and assess an automatic and robust knee musculoskeletal finite element (MSK-FE) modeling pipeline. <italic>Methods:</i> Magnetic resonance images (MRIs) were used to train nnU-Net networks for auto-segmentation of knee bones (femur, tibia, patella, and fibula), cartilages (femur, tibia, and patella), menisci, and major knee ligaments. Two different MRI sequences were used to broaden applicability. Next, we created MSK-FE models of an unseen dataset using two MSK-FE modeling pipelines: template-based and auto-meshing. MSK models had personalized knee geometries with multi-degree-of-freedom elastic foundation contacts. FE models used fibril-reinforced poroviscoelastic swelling material models for cartilages and menisci. <italic>Results:</i> Volumes of knee bones, cartilages, and menisci did not significantly differ (<italic>p</i>>0.05) across MRI sequences. MSK models estimated secondary knee kinematics during passive knee flexion tests consistent with <italic>in vivo</i> and simulation-based values from the literature. Between the template-based and auto-meshing FE models, estimated cartilage mechanics often differed significantly (<italic>p</i><0.05),>Conclusion:</i> The template-based modeling provided a more rapid and robust tool than the auto-meshing approach, while the estimated knee biomechanics were comparable. Nonetheless, the auto-meshing approach might provide more accurate estimates in subjects with distinct knee irregularities, e.g., cartilage lesions. <italic>Significance:</i> The MSK-FE modeling tool provides a rapid, easy-to-use, and robust approach for investigating task- and person-specific mechanical responses of the knee cartilage and menisci, holding significant promise, e.g., in personalized rehabilitation planning.","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"72 1","pages":"56-69"},"PeriodicalIF":4.4,"publicationDate":"2024-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10666719","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142139931","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：